Arctic Sea Ice Decline: Observations, Projections, Mechanisms, and Implications

Published by the American Geophysical Union as part of the Geophysical Monograph Series, Volume 180.

This volume addresses the rapid decline of Arctic sea ice, placing recent sea ice decline in the context of past observations, climate model simulations and projections, and simple models of the climate sensitivity of sea ice. Highlights of the work presented here include

• An appraisal of the role played by wind forcing in driving the decline;
• A reconstruction of Arctic sea ice conditions prior to human observations, based on proxy data from sediments;
• A modeling approach for assessing the impact of sea ice decline on polar bears, used as input to the U.S. Fish and Wildlife Service's decision to list the polar bear as a threatened species under the Endangered Species Act;
• Contrasting studies on the existence of a "tipping point," beyond which Arctic sea ice decline will become (or has already become) irreversible, including an examination of the role of the small ice cap instability in global warming simulations;
• A significant summertime atmospheric response to sea ice reduction in an atmospheric general circulation model, suggesting a positive feedback and the potential for short-term climate prediction.

The book will be of interest to researchers attempting to understand the recent behavior of Arctic sea ice, model projections of future sea ice loss, and the consequences of sea ice loss for the natural and human systems of the Arctic.

• Language: English
• Publisher: Wiley
• Release date: May 28, 2013
• ISBN: 9781118671580

Book preview

PREFACE

Prospects for Arctic sea ice are grim and apparently worsening. Following decades of decline, the September 2007 sea ice extent shattered all previous record lows with a 39% loss in ice cover relative to the 1979–2000 September mean. September 2007 also saw, for the first time on record, the opening of the northern branch of the Northwest Passage, the route through the Canadian Archipelago pioneered by Sir William Parry in 1819. These observations are accompanied by equally alarming climate model projections of sea ice decline due to greenhouse gas increases. Based on this evidence, expert predictions of the first appearance of an ice-free Arctic have advanced steadily.

The dramatic decline has prompted speculation in both the popular and scientific press that the Arctic sea ice may have passed a tipping point, beyond which the complete destruction of perennial sea ice cover is inevitable. While this claim may seem reasonable in light of the strong positive ice-albedo feedback, its validity has not been scientifically demonstrated. It does, however, suggest a set of questions regarding the climate sensitivity of Arctic sea ice: Will Arctic summer sea ice cover disappear completely? What factors control the rate of decline? Is there true threshold behavior? What can we learn about the climate sensitivity of Arctic sea ice from observations, both from the modern instrumented and paleo–proxy records? Are there meaningful idealized models that can be used to identify and isolate the dominant feedback mechanisms and assess climate sensitivity? Are there diagnostic analysis tools that can help quantify the declining trajectory of the sea ice? Can we quantify the climate impacts of sea ice decline?

This book is a collection of papers addressing these questions. It seeks to describe the climate sensitivity of Arctic sea ice in simplest terms, identifying the most prominent mechanisms and assessing their ability to produce a true, irreversible tipping point. The book is a combination of new results from original research and review material identifying key results from the literature that are insightful for understanding Arctic sea ice decline. The review material also serves to broaden the appeal of the book, which is intended for an audience including both specialists and interested nonspecialists with some background in physical climatology.

The book is divided into four sections. The first section, Arctic sea ice in the instrumented and paleo–proxy records sets the stage. Chapter 1 gives an assessment of sea ice cover change over recent decades, with emphasis on the contribution of wind forcing to the downward trend in sea ice cover. This discussion is followed in Chapter 2 by a reconstruction of sea ice conditions from the early Holocene, a period of reduced Arctic sea ice cover, using dinocyst assemblages and isotope ratios.

The second section, Factors in sea ice sensitivity considers mechanisms responsible for sea ice sensitivity in observations and climate model simulations. Chapter 3 considers the impact of clouds on the longwave and shortwave surface radiative fluxes over sea ice, and illustrates the contribution of cloudiness changes to sea ice loss in one climate change simulation. Chapter 4 considers the importance of the sea ice–albedo feedback as a source of spread in climate model simulations of sea ice decline, which is found to be small compared to the role of ice thickness differences already present at the start of these simulations. Chapter 5 continues the discussion of the large intermodel spread of Arctic sea ice thickness in climate model simulations, giving a simple argument for the intrinsic sensitivity of ice thickness to surface energy fluxes, and showing how differences in surface energy flux lead to the large spread in simulated sea ice thickness. Chapter 6 looks at a different kind of sensitivity: the sensitivity of an atmospheric circulation model subject to realistic reductions in summer Arctic sea ice.

The third section, Rapid loss versus abrupt transition, examines the extent to which true threshold behavior is implicated in climate model simulations that produce a seasonally ice-free Arctic in the 21st century. The notion of an abrupt, irreversible transition to ice-free Arctic summers, brought about by the destabilizing sea ice–albedo feedback, has become commonplace. But rapid loss does not necessarily imply a tipping point: strong sea ice sensitivity can combine with natural variability to produce rapid loss in the absence of a critical threshold between ice-free and ice-covered states. Chapter 7 considers the ability of the sea ice–albedo feedback to produce abrupt Arctic climate change, looking in particular at the role of the small ice cap instability in simulations in which Arctic sea ice vanishes in winter as well as summer. Chapter 8 analyzes simulations with a global climate model in which September Arctic sea ice vanishes as early as 2040, a rapid sea ice loss which apparently occurs through a combination of natural variability and anthropogenic forcing. Alternatively, chapter 9 presents a simple model in which abrupt loss can occur as a nonlinear transition between stable states after tuning to match one of the integrations described in the previous chapter. Chapter 10 quantifies the popular notion of the trajectory of Arctic sea ice, and shows trajectories of sea ice decline in a simplified phase space with dimensions representing ice cover of different thicknesses. A trajectory model in this phase space produces abrupt transitions when strong sea ice–albedo feedback is added. Chapter 11 gives an account of extreme loss events that occur as part of the natural variability of unforced climate model simulations. The characteristics of these unforced extreme events can shed light on the role of natural variability in forced sea ice decline.

Sea ice cover is a defining element of the Arctic ecosystem, and its loss will clearly have severe consequences for all life in the region. The final section of the book, The threat to polar bears from sea ice decline, describes a new modeling approach for assessing the impact of sea ice decline on the Arctic’s most iconic species. The model, presented in chapter 12, uses a Bayesian framework to determine the likely fate of polar bears in four distinct eco-regions of the Arctic, based in part on climate model projections of sea ice decline. The model was used to inform the recent decision by the Department of the Interior to list the polar bear as a threatened species under the Endangered Species Act.

This book grew out of a special session at the 2006 annual meeting of the American Geophysical Union titled Rapid transition from perennial to seasonal Arctic sea ice, in which many of the results in the chapters were presented. All chapters underwent a formal, anonymous review process. We gratefully acknowledge the efforts of the many individuals who served as reviewers for this monograph. Finally, we thank the chapter authors for their generous contributions, which made this book possible.

Eric T. DeWeaver

University of Wisconsin-Madison

Cecilia M. Bitz

University of Washington

L.-Bruno Tremblay

McGill University

Arctic Sea Ice Decline: Introduction

Eric T. DeWeaver

Center for Climate Research, University of Wisconsin-Madison, Madison, Wisconsin, USA

1. THE GREAT DECLINE OF 2007

By any measure, the loss of Arctic sea ice cover in September 2007 was spectacular. The National Snow and Ice Data Center (NSIDC) called it a loss the size of Alaska and Texas combined, in comparison to the 1979–2000 September mean. Record-breaking minima in sea ice extent are not unexpected, given the declining trend of the past 30 years and its recent acceleration [e.g., Meier et al., 2007; Deser and Teng, this volume]. But the 2007 minimum was remarkable even compared to the decline, a full four standard deviations below the trend line (H. Stern, quoted by Schweiger et al. [2008]). Kerr [2007] reported an Alaska-sized loss compared to the previous record low in 2005, which was itself an Alaska-sized retreat from the value at the beginning of the satellite era in 1979. Deser and Teng point out that the loss between September 2006 and September 2007 is as large as the entire September extent loss from 1979 to 2006.

Following the 2007 melt season there was some cause for optimism that 2008 could see a partial recovery. Writing at the end of the melt season, Comiso et al. [2008, paragraph 5] noted that the ice was rebounding with a rapid early autumn growth. Following a cold winter, the April 2008 maximum ice extent reported by NSIDC was relatively high by recent standards, although still below the long-term mean. But while the temperatures were cooperating, the winds were not. In early February I. Rigor noted that buoys embedded in multiyear ice flows were streaming out of the Arctic, flushed through Fram Strait along with their ice floes by circumpolar wind anomalies [Kizzia, 2008]. Also, as discussed by Maslanik et al. [2007], ice cover following the 2007 minimum was unusually thin and thus vulnerable to melting away. The July and August extent were somewhat higher than in 2007, but the daily loss rate accelerated in early August after storms broke apart thin ice in the Beaufort and Chukchi seas. Southerly winds following the storms further promoted opening by pushing the ice away from the eastern Siberian coast (information from the Arctic sea ice news and analysis Web pages for 11 August through 4 September 2008 at http://www.nsidc.org).

With approxmiately 2 weeks left in the 2008 melt season, Arctic sea ice extent is now very close to the 2007 minimum. While the lack of recovery is discouraging, the 2008 loss could have been worse. In May the NSIDC suggested, based on the prevalence of thin first year ice cover in the Arctic, that the North Pole could become ice free in 2008, a prediction more commonly made for the middle of the century. Three researchers contributing to the May Sea Ice Outlook (produced by the interagency Study of Environmental Arctic Change (SEARCH)), anticipated a return toward the long-term trend of summer sea ice loss, six argued that 2008 September extent should be close to 2007, and five expected losses exceeding those in 2007.

2. RESEARCH ON THE CAUSE OF THE LOSS

Research on the causes of the 2007 loss is already well underway. Surface wind anomalies are generally identified as the proximal cause [Nghiem et al., 2007; Stroeve et al., 2008; Deser and Teng, this volume; Overland et al., 2008; Zhang et al., 2008], as the transpolar winds dubbed the Polar Express by Nghiem et al. pushed ice away from the Alaskan and eastern Siberian coastlines and out of the Arctic. Kay et al. [2008] claim an additional role for enhanced summer melting as high pressure and sunny skies persisted over the western Arctic Ocean. Their claim is disputed by Schweiger et al. [2008], who note that the sunny skies are not well collocated with the largest sea ice loss in the Chukchi and East Siberian seas. On the other hand, a strong role for insolation as a positive feedback is not in dispute. Perovich et al. [2008] find a 500% increase in January to September solar heat input to the Beaufort Sea compared to the 1979—2005 climatology. They further determine that the large increase is due to the large area of low-albedo ocean surface exposed by the dramatic sea ice retreat. Accompanying the increased heat uptake, they report a sixfold increase in bottom melt measured by an ice mass buoy in the Beaufort Sea. Their results thus document the classical sea ice–albedo feedback, presumably initiated by wind-driven opening of the ice pack. The modeling study of Zhang et al. [2008] concludes that 70% of the 2007 loss anomaly was due to amplified melting while 30% resulted from ice motion.

In these studies, the meteorology of 2007 is generally given less prominence than the vulnerability of the 2007 sea ice cover. Maslanik et al. and Nghiem et al. document the long-term change from multiyear sea ice to younger floes which are thinner and more prone to breakup and melting. Overland et al. and Kay et al. also question the novelty of the 2007 meteorological conditions. They relate the offshore winds and sunny skies of 2007 to a surface high over the western Arctic Ocean, a rare but not unprecedented occurrence. Four years with comparable high pressure can be seen in the 50-year record shown by Overland et al. (their Figure 11), while Kay et al. find four additional years with sunnier skies than 2007. The older, thicker ice in these earlier years was not dramatically affected by the adverse meteorology.

The contribution of greenhouse warming in producing sharp, single-year declines is not easily quantified, since warming favors these events indirectly as it helps precondition the ice to a thinner state (e.g., Overland et al.). However, Stroeve et al. [2007] point out the consistency of the 2007 event with the periods of rapid loss found by Holland et al. [2006] in global warming simulations. Stroeve et al. note in particular the similarity between the March 2007 thickness estimates of Maslanik et al. and the mean Arctic thickness in simulations analyzed by Holland et al. The analysis of Holland et al. is expressed in terms of a three-part conceptual framework in which ice is first preconditioned for rapid loss by decades of thinning, after which loss is triggered by natural variability and then amplified by the sea ice–albedo feedback. The preconditioning, trigger, feedback framework was developed by Lindsay and Zhang [2005] to account for the observed sea ice decline from 1988 to 2003, and the same framework was invoked in the Zhang et al. study of the 2007 event. Thus, while the 2007 loss was unprecedented, descriptions of it are quite consistent with descriptions of the longer-term Arctic losses of the recent past and the rapid declines found in simulations of future Arctic change.

3. PAST THE TIPPING POINT?

Has the Arctic sea ice passed a tipping point? This is perhaps the most consistently asked question in news accounts about the 2007 and 2008 losses. Understandably, respondents to the question have not voiced much hope for reversal: "It’s hard to see how the system may come back (I. Rigor, quoted by Kizzia [2008]); I’m much more open to the idea that we might have passed a point where it’s becoming essentially irreversible" (J. M. Wallace, quoted by Revkin [2007]); It’s tipping now. We’re seeing it happen now (M. Serreze quoted by Borenstein and Joling [2008]). No doubt, the tipping point terminology aptly captures the precipitous loss of 2007 and lack of recovery in 2008. But questions remain as to how literally the tipping language should be taken. In a formal sense, tipping refers to a sudden and irreversible transition between two stable states of a system (e.g., right side up versus overturned), occurring as the system crosses some threshold value of a key parameter (e.g., angle to the local vertical). The scientific challenge would then be to find and characterize the stable states and threshold values of the Arctic sea ice system.

The idea of an unstable transition between ice-covered and ice-free Arctic states has a long history (see references of Winton and Merryfield et al. [this volume]), and such behavior does occur in simple energy balance models with diffusive heat transfer (the small ice cap instability of North [1984]). However, unstable transitions are somewhat elusive in global climate models, as Winton [this volume] shows. The rapid loss events in simulations of the Community Climate System Model (CCSM) shown by Holland et al. [2006] are commonly compared to the recent Arctic losses, yet threshold values for sea ice cover and thickness were not found for the CCSM events. Instead, the authors argue that rapid loss can occur through the superposition of natural variability and a steady downward trend (further analysis of these simulations is given by Holland et al. [this volume], Merrifield et al. [this volume], Stern et al. [this volume], and Gorodetskaya and Tremblay [this volume]. The lack of identifiable thresholds in CCSM is significant, since the yearly sea ice losses during CCSM rapid declines are larger than the 2007 loss observed by Holland et al., despite the absence of easily identifiable tipping points.

The primary motivation for claims of a tipping point comes from the destabilizing effect of the sea ice–albedo feedback. No doubt this is a strong feedback, but there is some subtlety in assessing its strength. Gorodetskaya and Tremblay point out that the effect of sea ice removal is mitigated by the cloudiness of the Arctic in summer, and note that the presence of sea ice reduces the top-of-atmosphere albedo by only 10 to 20%, despite the large albedo contrast between ice and open water. This finding is consistent with Winton’s [2006] conclusion that the sea ice albedo feedback is not dominant as a cause of polar amplification in climate models.

Moreover, the stability of the Arctic sea ice cover depends on the sign of the net feedback, with instability occurring when the positive sea ice–albedo feedback overwhelms the negative feedbacks which stabilize sea ice cover under colder conditions. Bitz [this volume] performed CCSM experiments in which Arctic Ocean surface albedo is held fixed even when sea ice cover is reduced by greenhouse gas increases, so that sea ice–albedo feedback is effectively disabled. The sea ice decline which occurs in the absence of sea ice–albedo feedback is not dramatically different from the sea ice decline in the control run. An explanation for this result is given by Winton [this volume], who performed model experiments in which sea ice cover was artificially removed. In these experiments increases in solar absorption due to increased open water area are offset by increases in turbulent heat flux from the ocean because of the removal of the insulating ice cover. The implication of these results is that the net feedback due to opening can still be negative, despite the strong positive sea ice–albedo feedback. Further support for this conclusion (at least in climate models) comes from Cullather and Tremblay’s [this volume] analysis of naturally occurring sea ice loss anomalies in a long CCSM control run with 1990 levels of greenhouse gases. Despite the sea ice–albedo feedback, sea ice cover rebounded within 1 to 3 years of each anomaly.

4. CLIMATE IMPACTS: POLAR BEAR LISTING DECISION

Of course, the implications of rapid sea ice loss go well beyond academic interest in climate stability. Policy makers are particularly challenged by Arctic sea decline, since they must plan for future sea ice conditions which are without precedent in the instrumented record. Faced with the lack of observed analogs, policy makers can seek guidance from global climate model (GCM) simulations of anthropogenic greenhouse warming. Such guidance can be quite valuable provided that two essential issues are addressed: first, the policy-relevant climate impacts of the simulated sea ice decline must be determined and, second, the uncertainty inherent in GCM projections of sea ice loss must be adequately assessed and incorporated. An important case in point is the research conducted by the U.S. Geological Survey (USGS) to advise the U.S. Fish and Wildlife Service (USFWS) on the impact of sea ice decline on polar bears. The research, which was presented in nine USGS administrative reports (online at www.usgs.gov/newsroom/special/polar_bears), was comissioned to help the USFWS decide whether to list the polar bear as a threatened species under the Endangered Species Act (ESA). Coincidentally, the results of this were presented to the USFWS in September 2007, as the Arctic sea ice cover approached its record low.

It is clear even upon superficial consideration that sea ice decline is bad for polar bears, given their dependence on sea ice as a platform for hunting and other activities (see references of Amstrup et al. [this volume]). However, the threat to polar bears from sea ice decline cannot be rigorously assessed without an understanding, based on observational field biology, of the sea ice needs of polar bears. Durner et al. [2008] quantified the habitat value of sea ice using observations of radio-collared polar bears over 2 decades. The characteristics that make sea ice desirable as polar bear habitat could be identified and quantified based on this data. In particular, polar bears were found to prefer sea ice over the shallow, productive waters of the continental shelf. The decline of pan-Arctic sea ice extent matters less than the retreat of sea ice from the shelf areas, as the habitat value of ice remaining over the deep Arctic basin is low.

Durner et al.’s resource selection functions (RSFs) quantify the value of sea ice as polar bear habitat, expressed as the frequency of occupation by polar bears, in terms of simple parameters including distance to shore, ocean depth, and sea ice concentration. The RSF methodology can be applied with equal ease to sea ice decline in observations and climate model projections. Thus, they enable researchers to provide guidance to policy makers in terms of the policy-relevant impact, in this case the loss of polar bear habitat, rather than generic statements regarding the overall sea ice decline. Further use of field data combined with model projections in the USGS reports comes from Hunter et al. [2007] who used data from a capture-release study to estimate declines in polar bear population as a function of reductions in sea ice availability.

Projections of future sea ice loss and its impacts will inevitably be accompanied by substantial uncertainty, given the evident sensitivity of the Arctic climate system. As discussed by Amstrup et al. [this volume], the USGS research accounted for model uncertainty by using a subset of 10 climate models which satisfy a selection criterion based on present day sea ice simulation quality. Projections from this subset show a range of September sea ice loss from 30 % to complete loss by mid century (sources of uncertainty in sea ice projections are discussed by Bitz and DeWeaver et al. [this volume]). The uncertainty represented by the range of model simulations was propagated through the USGS analysis by applying techniques like the RSF calculation to the whole subset, so that ensemble spread in sea ice simulations leads to ensemble spread in polar bear outcomes. However, Amstrup et al. note that these projections may be overly optimistic, given Stroeve et al.’s [2007] finding that real-world Arctic sea ice has declined at almost twice the rate found in model simulations of the recent past. The USGS efforts culminated in Amstrup et al.’s synthesis report, which uses a Bayesian framework to assess the probability of decline in polar bear population based on consideration of sea ice decline and other factors. Despite the uncertainties of the research, none of the outcomes were favorable for polar bears; in effect, they run the gamut from bad to extremely bad.

In May 2008 the polar bear was listed as a threatened species under ESA, after considerable delay. It is clear from the final rule [U.S. Fish and Wildlife Service, 2008] that the policy makers understood and considered the scientific guidance. Consideration of the science is also evident from the announcement of the decision (www.doi.gov/secretary/speeches/081405_speech.html), which included a prominent display of Stroeve et al.’s [200X] work on observed and simulated sea ice trends, and maps of Arctic sea ice showing the change in coverage by old (at least 5 years) and new (less than 5 years old) ice, apparently from the drift model of Rigor and Wallace [2004]. But while the effort to provide scientific input for the listing decision was successful in some sense, it remains to be seen if the listing will have any direct effect on the status of the polar bear (see analysis of Revkin [2008]).

5. CONCLUSION

The events of 2007 and 2008 highlight the need for improved understanding of sea ice sensitivity and the impacts of sea ice decline. Perhaps, if we are fortunate, our understanding of the Arctic sea ice and climate system can evolve fast enough to keep pace with the changes occurring there.

Acknowledgments. The author’s research is supported by the Office of Science (BER), U.S. Department of Energy, grant DE-FG02-03ER63604. I thank Cecilia Bitz, Steven Amstrup and members of the NCAR Polar Climate Working Group for helpful conversations. I am indebted to the chapter authors for contributing their best work to this monograph.

REFERENCES

Amstrup, S. C., B. G. Marcot, and D. C. Douglas (2008), A Bayesian network model approach to forecasting the 21st century worldwide status of polar bears, this volume.

Bitz, C. M. (2008), Some aspects of uncertainty in predicting sea ice thinning, this volume.

Borenstein, S., and D. Joling (2008), Arctic sea ice drops to 2nd lowest level on record, San Francisco Chron., 27 Aug.

Comiso, J. C., C. L. Parkinson, R. Gersten, and L. Stock (2008), Accelerated decline in the Arctic sea ice cover, Geophys. Res. Lett., 35, L01703, doi:10.1029/2007GL031972.

Deser, C., and H. Teng (2008), Recent trends in Arctic sea ice and the evolving role of atmospheric circulation forcing, 1979–2007, this volume.

de Vernal, A., C. Hillaire-Marcel, S. Solignac, T. Radi, and A. Rochon (2008), Reconstructing sea-ice conditions in the Arctic and subarctic prior to human observations, this volume.

DeWeaver, E. T., E. C. Hunke, and M. M. Holland (2008), Sensitivity of Arctic sea ice thickness to intermodel variations in the surface energy budget, this volume.

Durner, G. M., et al. (2008), Predicting 21st century polar bear habitat distribution from global climate models, Ecol. Monogr., in press.

Gorodetskaya, I. V., and L.-B. Tremblay (2008), Arctic cloud properties and radiative forcing from observations and their role in sea ice decline predicted by the NCAR CCSM3 model during the 21st century, this volume.

Holland, M. M., C. M. Bitz, and B. Tremblay (2006), Future abrupt reductions in the summer Arctic sea ice, Geophys. Res. Lett., 33, L23503, doi:10.1029/2006GL028024.

Holland, M. M., C. M. Bitz, B. Tremblay, and D. A. Bailey (2008), The role of natural versus forced change in future rapid summer Arctic ice loss, this volume.

Kay, J. E., T. L’Ecuyer, A. Gettelman, G. Stephens, and C. O’Dell (2008), The contribution of cloud and radiation anomalies to the 2007 Arctic sea ice extent minimum, Geophys. Res. Lett., 35, L08503, doi:10.1029/2008GL033451.

Kerr, R. A. (2007), Is battered sea ice down for the count?, Science, 318, 33–34.

Kizzia, T. (2008), Polar ice pack loss may break 2007 record, Anchorage Daily News, 12 Feb.

Lindsay, R. W., and J. Zhang (2005), The thinning of Arctic sea ice, 1988–2003: Have we passed a tipping point?, J. Clim., 18, 4879–4894.

Maslanik J. A., C. Fowler, J. Stroeve, S. Drobot, J. Zwally, D. Yi, and W. Emery (2007), A younger, thinner Arctic ice cover: Increased potential for rapid, extensive sea-ice loss, Geophys. Res. Lett., 34, L24501, doi:10.1029/2007GL032043.

Meier, W. N., J. Stroeve, and F. Fetterer (2007), Wither Arctic sea ice? A clear signal of decline, regionally, seasonally, and extending beyond the satellite record, Ann. Glaciol., 46, 428–434.

Merryfield, W. J., M. M. Holland, and A. H. Monahan (2008), Multiple equilibria and abrupt transitions in Arctic summer sea ice extent, this volume.

Nghiem, S. V., I. G. Rigor, D. K. Perovich, P. Clemente-Colón, J. W. Weatherly, and G. Neumann (2007), Rapid reduction of Arctic perennial sea ice, Geophys. Res. Lett., 34, L19504, doi:10.1029/2007GL031138.

North, G. R. (1984), The small ice cap instability in diffusive climate models, J. Atmos. Sci., 41, 3390–3395.

Overland, J. E., M. Wang, and S. Salo (2008), The recent Arctic warm period, Tellus, Ser. A, 60, 589–597.

Perovich, D. K., J. A. Richter-Menge, K. F. Jones, and B. Light (2008), Sunlight, water, and ice: Extreme Arctic sea ice melt during the summer of 2007, Geophys. Res. Lett., 35, L11501, doi:10.1029/2008GL034007.

Revkin, A. C. (2007), Arctic melt unnerves the experts, N. Y. Times, 2 Oct.

Revkin, A. C. (2008), Polar bear is made a protected species, N. Y. Times, 15 May.

Rigor, I. G., and J. M. Wallace (2004), Variations in the age of Arctic sea-ice and summer sea-ice extent, Geophys. Res. Lett., 31, L09401, doi:10.1029/2004GL019492.

Schweiger, A. J., J. Zhang, R. W. Lindsay, and M. Steele (2008), Did unusually sunny skies help drive the record sea ice minimum of 2007?, Geophys. Res. Lett., 35, L10503, doi:10.1029/2008GL033463.

Schweiger, A. J., J. Zhang, R. W. Lindsay, and M. Steele (2008), Did unusually sunny skies help drive the record sea ice minimum of 2007?, Geophys. Res. Lett., 35, L10503, doi:10.1029/2008GL033463

Stern, H. L., R. W. Lindsay, C. M. Bitz, and P. Hezel (2008), What is the trajectory of Arctic sea ice?, this volume.

Stroeve J., M. M. Holland, W. Meier, T. Scambos, and M. Serreze (2007), Arctic sea ice decline: Faster than forecast, Geophys. Res. Lett., 34, L09501, doi:10.1029/2007GL029703.

Stroeve, J., M. Serreze, S. Drobot, S. Gearheard, M. Holland, J. Maslanik, W. Meier, and T. Scambos (2008), Arctic sea ice extent plummets in 2007, Eos Trans. AGU, 89(2), 13.

U.S. Fish and Wildlife Service (2008), Endangered and threatened wildlife and plants; determination of threatened status for the polar bear (Ursus maritimus) throughout its range, Fed. Regist., 73, 28,211–28,303.

Winton, M. (2006), Amplified Arctic climate change: What does surface albedo feedback have to do with it?, Geophys. Res. Lett., 33, L03701, doi:10.1029/2005GL025244.

Winton, M. (2008), Sea ice–albedo feedback and nonlinear Arctic climate change, this volume.

Zhang, J., R. Lindsay, M. Steele, and A. Schweiger (2008), What drove the dramatic retreat of Arctic sea ice during the summer 2007?, Geophys. Res. Lett., 35, L11505, doi:10.1029/ 2008GL034005.

E. T. DeWeaver, Center for Climate Research, University of Wisconsin-Madison, 1225 West Dayton Street, Madison, WI 53706, USA. ([e-mail])

Section I

Arctic Sea Ice in the Instrumented and Paleo–Proxy Records

Recent Trends in Arctic Sea Ice and the Evolving Role of Atmospheric Circulation Forcing, 1979–2007

Clara Deser and Haiyan Teng

National Center for Atmospheric Research, Boulder, Colorado, USA

This study documents the evolving trends in Arctic sea ice extent and concentration during 1979–2007 and places them within the context of overlying changes in the atmospheric circulation. Results are based on 5-day running mean sea ice concentrations (SIC) from passive microwave measurements during January 1979 to October 2007. Arctic sea ice extent has retreated at all times of the year, with the largest declines (0.65 × 10⁶ km² per decade, equivalent to 10% per decade in relative terms) from mid July to mid October. The pace of retreat has accelerated nearly threefold from the first half of the record to the second half, and the number of days with SIC less than 50% has increased by 19 since 1979. The spatial patterns of the SIC trends in the two halves of the record are distinctive, with regionally opposing trends in the first half and uniformly negative trends in the second half. In each season, these distinctive patterns correspond to the first two leading empirical orthogonal functions of SIC anomalies during 1979–2007. Atmospheric circulation trends and accompanying changes in wind-driven atmospheric thermal advection have contributed to thermodynamic forcing of the SIC trends in all seasons during the first half of the record and to those in fall and winter during the second half. Atmospheric circulation trends are weak over the record as a whole, suggesting that the long-term retreat of Arctic sea ice since 1979 in all seasons is due to factors other than wind-driven atmospheric thermal advection.

1. INTRODUCTION

The accelerating retreat of Arctic sea ice in recent decades, evident in all months of the year, is one of the most dramatic signals of climate change worldwide (see Serreze et al. [2007], Meier et al. [2007], and Stroeve et al. [2007] for recent overviews; ongoing updates on Arctic sea ice may be obtained from the National Snow and Ice Data Center (available at http://nsidc.org)). Although climate models predict that Arctic sea ice will decline in response to atmospheric greenhouse gas increases [Holland et al., 2006], the current pace of retreat at the end of the melt season is exceeding the models’ forecasts by approximately a factor of 3 [Stroeve et al., 2007]. Long-term records of summer sea ice extent within the central Arctic Ocean dating back to 1900 exhibit large multidecadal variations [Polyakov et al., 2003], a factor which must be taken into account when interpreting the recent sea ice retreat.

The physical mechanisms underlying the Arctic sea ice decline are not fully understood but include dynamical processes related to changes in winds and ocean currents and thermodynamic processes involving changes in air temperature, radiative and turbulent energy fluxes, ocean heat storage, and ice-albedo feedback [Serreze et al., 2007; Stroeve and Maslowski, 2008; Francis and Hunter, 2006, 2007; Shimada et al., 2006; Perovich et al., 2007]. A better understanding of these mechanisms and their relationship to increasing greenhouse gas concentrations is an important step for assessing future predictions of Arctic climate change.

Numerous studies indicate that the atmospheric circulation played an important role in driving Arctic sea ice declines from the 1960s to the early 1990s [e.g., Deser et al., 2000; Rigor et al., 2002; Hu et al., 2002; Rigor and Wallace, 2004; Rothrock and Zhang, 2005; Stroeve et al., 2007; Serreze and Francis, 2006; Ukita et al., 2007]. In particular, the declines during this period were due in part to a trend in the dominant pattern of wintertime atmospheric circulation variability over the high-latitude Northern Hemisphere known variously as the North Atlantic Oscillation, Arctic Oscillation, or Northern Annular Mode [Hurrell, 1995; Deser, 2000; Thompson and Wallace, 2000], collectively referred to hereinafter as the NAM. In particular, the anomalous cyclonic wind circulation associated with the upward trend in the winter NAM flushed old, thick ice out of the Arctic via Fram Strait, causing the winter ice pack to thin, which, in turn, preconditioned the summer ice pack for enhanced melt.

Since the early 1990s, however, the trend in the NAM has reversed sign, yet Arctic sea ice has continued to decline [Overland and Wang, 2005; Comiso, 2006; Serreze and Francis, 2006; Maslanik et al., 2007; Serreze et al., 2007; Stroeve and Maslowski, 2008]. This has led to speculation that the Arctic climate system has reached a tipping point whereby strong positive feedback mechanisms such as those associated with ice albedo and open water formation efficiency are accelerating the thinning and retreat of Arctic sea ice [e.g., Lindsay and Zhang, 2005; Holland et al., 2006]. These positive feedback mechanisms leave the ice pack more vulnerable to forcing from other processes, natural and anthropogenic. For example, enhanced downward longwave radiation associated with increases in air temperature, water vapor and cloudiness over the Arctic Ocean [Francis and Hunter, 2006] along with enhanced ocean heat transport into the Arctic [Polyakov et al., 2005; Shimada et al., 2006; Stroeve and Maslowski, 2008] and positive ice-albedo feedback [Perovich et al., 2007] have become dominant factors driving summer sea ice extent declines since the mid-to-late 1990s. There is also evidence that the winter atmospheric circulation has continued to affect the winter sea ice distribution since the mid-1990s [Comiso, 2006; Maslanik et al., 2007; Francis and Hunter, 2007].

The purpose of this study is to revisit the issue of Arctic sea ice trends from 1979 to present in the context of evolving atmospheric circulation conditions. In addition to examining trends over the entire period of record, we investigate trends over the two halves separately as a simple way of characterizing their evolution. We note that the first half coincides with an upward trend in the NAM, while the second half coincides with a downward trend. We are particularly interested in assessing the evolving role of thermodynamic atmospheric circulation forcing of sea ice concentration trends, taking into account any seasonal dependencies. We use 5-day running mean sea ice concentration data on a 25 km × 25 km grid derived from passive microwave measurements from 1 January 1979 through 31 October 2007. Early results were presented by Deser and Teng [2008] for the winter and summer seasons only.

Our study is organized as follows. Section 2 describes the data sets and methodology. Section 3.1 provides results on trends in Arctic sea ice extent throughout the annual cycle, as well as derived quantities such as the timing of the seasonal cycle. Section 3.2 presents the spatial patterns of sea ice concentration, sea level pressure, and wind-induced atmospheric thermal advection trends for the two halves of the study period and for the record as a whole, stratified by season. Air temperature, sea surface temperature (SST), and net surface downward longwave radiation trends from 1979 to present are also shown. Section 4 provides a summary and discussion of the results.

2. DATA AND METHODS

Daily sea ice concentrations (SIC) on a 25 km × 25 km grid for the period 1 January 1979 to 31 October 2007 were obtained from the National Snow and Ice Data Center. These data are derived from the Nimbus 7 Scanning Multichannel Microwave Radiometer and Defense Meteorological Satellite Program (DMSP) F8, F11, and F13 Special Sensor Microwave/Imager radiances using the NASA team algorithm [Cavalieri et al., 1999].

In addition to daily SIC, we use monthly mean sea level pressure (SLP), 1000 hPa zonal and meridional wind components, 2-m and 1000-hPa air temperatures on a 2.5° × 2.5° latitude grid from the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) reanalysis project [Kalnay et al., 1996] for the period January 1979 through October 2007. We also use monthly mean sea surface temperature data from the Had-ISST1 data set [Rayner et al., 2003] on a 1° × 1 ° latitude grid, updated through December 2006. The NCEP/NCAR reanalysis and HadISST1 data sets were obtained from the Data Support Section at NCAR. Finally, we make use of daily net surface downward longwave radiative fluxes derived from the NASA-NOAA Television and Infrared Observation Satellite (TIROS) Operational Vertical Sounder (TOVS) polar pathfinder data set [Francis and Hunter, 2007]. These data are available from July 1979 through December 2005 on a 100-km² grid north of 55°N.

3. RESULTS

3.1. Arctic Sea Ice Extent

A map of the locations of the Arctic and sub-Arctic seas referred to in this study are shown in Figure 1. These place names are superimposed upon the long-term mean distributions of maximum and minimum sea ice extent (defined as marine areas within which sea ice concentrations equal or exceed 15%), based on the period 1979–2006. The maximum sea ice extent is defined as the 30-day average centered on the mean date of maximum extent, 7 March, and the minimum extent is defined as the 30-day average centered on the mean date of minimum extent, 17 September. At maximum extent, all of the Arctic and sub-Arctic seas are ice covered (sea ice concentrations >15%), while at minimum extent, only the Greenland and Beaufort seas and the Arctic basin (central Arctic Ocean) are ice covered.

Figure 1. Locations of the Arctic and sub-Arctic seas referred to in this study, superimposed upon the long-term (1979–2007) mean sea ice extent at month of maximum (thick black contour) and month of minimum (thin black contour and light shaded areas). The maximum (minimum) sea ice extent is defined as the 30-day average centered on the mean date of maximum (minimum) extent, 7 March (17 September).

c02_image001.jpg

Arctic sea ice extent serves as a useful starting point for describing the temporal character of sea ice over the Arctic as a whole. Following convention, we have defined Arctic sea ice extent as the area of the ocean covered by at least 15% sea ice concentration based on 5-day